This is a newly installed server running CentOS 4 with all current updates. It's been cruising nicely since installation a week ago until this morning. Then everyting started timing out and the attached messages appeared in the log (this is a short and sanitized extract). Load went sky high.
Had to reboot to restore service and all looks happy again now, but if anyone can provide some input on these messages...
TIA Brian
Brian Parish wrote:
This is a newly installed server running CentOS 4 with all current updates. It's been cruising nicely since installation a week ago until this morning. Then everyting started timing out and the attached messages appeared in the log (this is a short and sanitized extract). Load went sky high.
Had to reboot to restore service and all looks happy again now, but if anyone can provide some input on these messages...
TIA Brian
Maybe you should look at your sar reports. In /var/log/sa
On Thu, 2005-06-16 at 11:07 +1000, Brian Parish wrote:
This is a newly installed server running CentOS 4 with all current updates. It's been cruising nicely since installation a week ago until this morning. Then everyting started timing out and the attached messages appeared in the log (this is a short and sanitized extract). Load went sky high.
Had to reboot to restore service and all looks happy again now, but if anyone can provide some input on these messages...
TIA Brian
There are known (upstream :) issues with memory leaks on the 2.6.9-5.0.3 and the 2.6.9-5.0.5 kernels. Those are supposed to be fixed on the 2.6.9-11 kernel (which you should have if you are updating your kernel).
On Friday 17 June 2005 07:22, Johnny Hughes wrote:
On Thu, 2005-06-16 at 11:07 +1000, Brian Parish wrote:
This is a newly installed server running CentOS 4 with all current updates. It's been cruising nicely since installation a week ago until this morning. Then everyting started timing out and the attached messages appeared in the log (this is a short and sanitized extract). Load went sky high.
Had to reboot to restore service and all looks happy again now, but if anyone can provide some input on these messages...
TIA Brian
There are known (upstream :) issues with memory leaks on the 2.6.9-5.0.3 and the 2.6.9-5.0.5 kernels. Those are supposed to be fixed on the 2.6.9-11 kernel (which you should have if you are updating your kernel).
This could be it - it was still running 2.6.9-5.0.5.EL - now updating. Lazy sysadmin!
thanks Brian
-
Brian Parish -
Johnny Hughes -
Ugo Bellavance